On quantifying passing skill

Quantifying passing skill has been a topic that has gained greater attention over the past 18 months in public analytics circles, with Paul Riley,  StatsBomb and Played off the Park regularly publishing insights from their passing models. I talked a little about my own model last season but only published results on how teams disrupted their opponents passing. I thought delving into the nuts and bolts of the model plus reporting some player-centric results would be a good place to start as I plan to write more on passing over the next few months.

Firstly, the model quantifies the difficulty of an open-play pass based on its start and end location, as well as whether it was with the foot or head. So for example, relatively short backward passes by a centre back to their goalkeeper are completed close to 100% of the time, whereas medium-range forward passes from out-wide into the centre of the penalty area have pass completion rates of around 20%.

The data used for the model is split into training and testing sets to prevent over-fitting. The Random Forest-based model does a pretty good job of representing the different components that drive pass difficulty, some of which are illustrated in the figure below (also see the appendix here for some further diagnostics).

xP_val

Comparison between expected pass completion rates from two different passing models and actual pass completion rates based on the start and end location of an open-play pass. The horizontal dimension is orientated from left-to-right, with zero designating the centre of the pitch. The dashed lines in the vertical dimension plots show the location of the edge of each penalty area. Data via Opta.

One slight wrinkle with the model is that it has trouble with very short passes of less than approximately 5 yards due to the way the data is collected; if a player attempts a pass and an opponent in his immediate vicinity blocks it, then the pass is unsuccessful and makes it looks like such passes are really hard, even though the player was actually attempting a much longer pass. Neil Charles reported something similar in his OptaPro Forum presentation in 2017. For the rest of the analysis, such passes are excluded.

None shall pass

That gets some of the under-the-hood stuff out of the way, so let’s take a look at ways of quantifying passing ‘skill’.

Similar to the concept of expected goals, the passing model provides a numerical likelihood of a given pass being completed by an average player; deviations from this expectation in reality may point to players with greater or less ‘skill’ at passing. The analogous concept from expected goals would be comparing the number of goals scored versus expectation and interpreting this as ‘finishing skill‘ or lack there of. However, when it comes to goal-scoring, such interpretations tend to be very uncertain due to significant sample size issues because shots and goals are relatively infrequent occurrences. This is less of a concern when it comes to passing though, as many players will often attempt more passes in a season than they would take shots in their entire career.

Another basic output of such models is an indication of how adventurous a player is in their passing – are they playing lots of simple sideways passes or are they regularly attempting defense-splitting passes?

The figure below gives a broad overview of these concepts for out-field players from the top-five leagues (England, France, Germany, Italy and Spain) over the past two seasons. Only passes with the feet are included in the analysis.

dxP_avg_xP_scatter

Passing ‘skill’ compared to pass difficulty for outfield players from the past two seasons in the big-five leagues, with each data point representing a player who played more than 3420 minutes (equivalent to 38 matches) over the period. The dashed lines indicate the average values across each position. Foot-passes only. Data from Opta.

One of the things that is clear when examining the data is that pulling things apart by position is important as the model misses some contextual factors and player roles obviously vary a huge amount depending on their position. The points in the figure are coloured according to basic position profiles (I could be more nuanced here but I’ll keep it simpler for now), with the dashed lines showing the averages for each position.

In terms of pass difficulty, midfielders attempt the easiest passes with an average expected completion of 83.2%. Forwards (81.6%) attempt slightly easier passes than defenders (81.4%), which makes sense to me when compared to midfielders, as the former are often going for tough passes in the final third, while the latter are playing more long passes and crosses.

Looking at passing skill is interesting, as it suggest that the average defender is actually more skilled than the average midfielder?!? While the modern game requires defenders to be adept in possession, I’m unconvinced that their passing skills outstrip midfielders. What I suspect is happening is that passes by defenders are being rated as slightly harder than they are in reality due to the model not knowing about defensive pressure, which on average will be less for defenders than midfielders.

Forwards are rated worst in terms of passing skill, which is probably again a function of the lack of defensive pressure included as a variable, as well as other skills being more-valued for forwards than passing e.g. goal-scoring, dribbling, aerial-ability.

Pass muster

Now we’ve got all that out of the way, here are some lists separated by position. I don’t watch anywhere near as much football as I once did, so really can’t comment on quite a few of these and am open to feedback.

Note the differences between the players on these top-ten lists are tiny, so the order is pretty arbitrary and there are lots of other players that the model thinks are great passers who just missed the cut.

First-up, defenders: *shrugs*.

In terms of how I would frame this, I wouldn’t say ‘Faouzi Ghoulam is the best passer out of defenders in the big-five leagues’. Instead I would go for something along the lines of ‘Faouzi Ghoulam’s passing stands out and he is among the best left-backs according to the model’. The latter is more consistent with how football is talked about in a ‘normal’ environment, while also being a more faithful presentation of the model.

Looking at the whole list, there is quite a range of pass difficulty, with full-backs tending to play more difficult passes (passes into the final third, crosses into the penalty area) and the model clearly rates good-crossers like Ghoulam, Baines and Valencia. Obviously that is a very different skill-set to what you would look for in a centre back, so filtering the data more finely is an obvious next step.

Defenders (* denotes harder than average passes)

Name Team xP rating Pass difficulty
Faouzi Ghoulam Napoli 1.06 80.3*
Leighton Baines Everton 1.06 76.5*
Stefan Radu Lazio 1.06 82.1
Thiago Silva PSG 1.06 91.0
Benjamin Hübner Hoffenheim 1.05 84.4
Mats Hummels Bayern Munich 1.05 86.0
Kevin Vogt Hoffenheim 1.05 87.4
César Azpilicueta Chelsea 1.05 83.4
Kalidou Koulibaly Napoli 1.05 87.8
Antonio Valencia Manchester United 1.05 80.0*

On to midfielders: I think this looks pretty reasonable with some well-known gifted passers making up the list, although I’m a little dubious about Dembélé and Fernandinho being quite this high up. Iwobi is an interesting one and will keep James Yorke happy.

Fàbregas stands-out due to his pass difficulty being well-below average without having a cross-heavy profile – nobody gets near him for the volume of difficult passes he completes.

Midfielders (* denotes harder than average passes)

Name Team xP rating Pass difficulty
Cesc Fàbregas Chelsea 1.06 79.8*
Toni Kroos Real Madrid 1.06 88.1
Luka Modric Real Madrid 1.06 85.9
Arjen Robben Bayern Munich 1.05 79.6*
Jorginho Napoli 1.05 86.8
Mousa Dembélé Tottenham Hotspur 1.05 89.9
Fernandinho Manchester City 1.05 87.2
Marco Verratti PSG 1.05 87.3
Alex Iwobi Arsenal 1.05 84.9
Juan Mata Manchester United 1.05 84.5

Finally, forwards AKA ‘phew, it thinks Messi is amazing’.

Özil is the highest-rated player across the dataset, which is driven by his ability to retain possession and create in the final third. Like Fàbregas above, Messi stands out for the difficulty of the passes he attempts and that he is operating in the congested central and half-spaces in the final third, where mere mortals (and the model) tend to struggle.

In terms of surprising names: Alejandro Gomez appears to be very good at crossing, while City’s meep-meep wide forwards being so far up the list makes we wonder about team-effects.

Also, I miss Philippe Coutinho.

Forwards (* denotes harder than average passes)

Name Team xP rating Pass difficulty
Mesut Özil Arsenal 1.07 82.9
Eden Hazard Chelsea 1.05 81.9
Lionel Messi Barcelona 1.05 79.4*
Philippe Coutinho Liverpool 1.04 80.6*
Paulo Dybala Juventus 1.03 84.8
Alejandro Gomez Atalanta 1.03 74.4*
Raheem Sterling Manchester City 1.03 81.6*
Leroy Sané Manchester City 1.03 81.9
Lorenzo Insigne Napoli 1.03 84.3
Diego Perotti Roma 1.02 78.4*

Finally, the answer to what everyone really wants to know is, who is the worst passer? Step-forward Mario Gómez – I guess he made the right call when he pitched his tent in the heart of the penalty area.

Pass it on

While this kind of analysis can’t replace detailed video and live scouting for an individual, I think it can provide a lot of value. Traditional methods can’t watch every pass by every player across a league but data like this can. However, there is certaintly a lot of room for improvement and further analysis.

A few things I particularly want to work on are:

  • Currently there is no information in the model about the type of attacking move that is taking place, which could clearly influence pass difficulty e.g. a pass during a counter-attacking situation or one within a long passing-chain with much slower build-up. Even if you didn’t include such parameters in the model, it would be a nice means of filtering different pass situations.
  • Another element in terms of context is attempting a pass after a dribble, especially given some of the ratings above e.g. Hazard and Dembélé. I can envisage the model somewhat conflates the ability to create space through dribbling and passing skill (although this isn’t necessarily a bad thing depending on what you want to assess).
  • Average difficulty is a bit of a blunt metric and hides a lot of information. Developing this area should be a priority for more detailed analysis as I think building a profile of a player’s passing tendencies would be a powerful tool.
  • You’ll have probably noticed the absence of goalkeepers in the results above. I’ve left them alone for now as the analysis tends to assign very high skill levels to some goalkeepers, especially those attempting lots of long passes. My suspicion is that long balls up-field that are successfully headed by a goalkeeper’s team-mate are receiving a bit too much credit i.e. yes the pass was ‘successful’ but that doesn’t always mean that possession was retained after the initial header. That isn’t necessarily the fault of the goalkeeper, who is generally adhering to the tactics of their team and the match situation but I’m not sure it really reflects what we envisage as passing ‘skill’ when it comes to goalkeepers. Discriminating between passes to feet and aerial balls would be a useful addition to the analysis here.
  • Using minutes as the cut-off for the skill ratings leaves a lot of information on the table. The best and worst passers can be pretty reliably separated after just a few hundred passes e.g. Ruben Loftus-Cheek shows up as an excellent passer after just ~2000 minutes in the Premier League. Being able to quickly assess young players and new signings should be possible. Taking into account the number of passes a player makes should also be used to assess the uncertainty in the ratings.

I’ve gone on enough about this, so I’ll finish by saying that any feedback on the analysis and ratings is welcome. To facilitate that, I’ve built a Tableau dashboard that you can mess around with that is available from here and you can find the raw data here.

Time to pass and move.

Advertisements

Pressing Issues at the World Cup

One of the tactical questions heading into the World Cup was whether the modern popularity of pressing at club level would be replicated on the international stage. The best pressing teams are a meld of intelligent positioning and trigger movements, honed during hours on the training pitch, requiring intense and sustained athletic performance.

Such luxuries are not afforded to international teams, who have far less training time to drill such tactics, coupled with many players arriving at the tournament after long club seasons. A heatwave encompassing much of Europe, including many of the west Russian venues, presents another potential barrier to intense pressing at the tournament.

Gegenpressing or counter-pressing is one form of pressing, where a team presses the opposition immediately after losing possession, with the initial aim being to disrupt the opponent’s transition phase and potential counter-attack. Across the English Premier League last season, 46% of open-play possessions were pressed within 5 seconds, with the trio of Spurs, Liverpool and Manchester City leading the league with a counter-pressing fraction of close to 60%. At the other end of the scale was Stoke City and West Brom at around 35%.

Perhaps surprisingly, the counter-pressing rate at the World Cup through the group stage and last-sixteen stands at 59%. There are a number of contextual factors to these figures that partially explain this difference. Note the absolute figures below are per game per team.

Firstly there have been slightly fewer open-play possessions available to counter-press at the World Cup (33) than in the EPL (35); fewer open-play possessions potentially means that teams are able to maintain a higher intensity, which would shift the observed counter-pressing fractions. This is borne out in the average number of counter-pressed possessions, which stands at 19 possessions in the World Cup compared to 16 possessions in the EPL.

Secondly there is a significant stylistic contrast between the EPL and World Cup in terms of passing, with 66 long balls per game in the former and 59 per game in the latter. Conversely, short passes are greater (398 per game) at the World Cup than in the past EPL season (392). The greater propensity towards long balls in the EPL, especially early on in a possession, cuts down on opportunities to counter-press, while a lower number of shorter passes will act in a similar manner.

The third factor is the differing priorities and incentives of a World Cup group stage and a 38-game season. The figure below illustrates the importance of considering the small sample size at play in the World Cup and how a team’s approach will vary from game-to-game. Each team is ranked according to their average counter-pressing percentage, with their individual matches shown to illustrate the variation across the tournament. This variation is further summarized in the right-hand figure, where the standard deviation across their matches is shown as the horizontal lines around each data point (broader lines mean a greater standard deviation and thus more variation).

World-Cup-Counter-Pressing-Smaller

Iceland are perhaps the best example of how a group stage spanning three matches shapes tactical considerations and can lead to misleading averages – in their first match against Argentina, they registered the lowest counter-pressing percentage of the tournament as they were happy to get players behind the ball and concentrate on frustrating their opponent’s efforts in the final third. A draw against the perceived giant of the group was a good result on their World Cup debut but subsequent matches would require them to force the issue more in search of a win to place them in a good position to advance; their counter-pressing percentage increased in their subsequent matches culminating in a very high counter-pressing rate against Croatia with a win needed for qualification from the group stage.

Contrast this with a team in the EPL that is uninclined to counter-press, such as West Brom, when playing against more favoured opponents – three draws or perhaps a narrow fortuitous win would often be a more than satisfactory outcome, so sitting deep and frustrating the opposition is a sensible strategy. Such considerations apply at the other end of the scale also e.g. taking a small sample of matches against a minnow sitting deep in their own half from Spurs’ season when chasing a win would yield a selection of matches where they aggressively counter-pressed their opponent. This latter example likely explains the very high counter-pressing rate by the likes of Germany, Spain and Argentina as they found themselves in close matches where they needed to push for a win.

This isn’t to say that the tactical approach of all teams at the World Cup is dictated more by game-state and qualification permutations – of the teams remaining Sweden, Uruguay, Croatia and Belgium have been quite consistent across the tournament and are perhaps more likely to employ their preferred counter-pressing intensity in their remaining matches. On the other side of the coin, Brazil, France and England have been more variable with Russia sat in the middle.

Pressing outcomes

In terms of evaluating successful counter-pressing, the gap between the EPL and World Cup is less obvious – 17% of counter-pressing possessions have seen possession won or disrupted within 5 seconds in the World Cup, which is only 1% greater than in the EPL. In both the EPL and World Cup, 10% of counter-pressing possessions have ended with a shot for the team being pressed, which results in an increased absolute rate (1.9 shots per game) in the World Cup compared to the EPL (1.6 per game).

In the EPL, the most aggressive counter-pressing teams tend to also routinely win possession back and concede fewer shots during this phase (Manchester City top the rankings in both categories, with Liverpool not too far behind and Spurs performing at an above average level). Such figures are likely noisier during the World Cup given the small sample size with the trend being similar albeit less strong, with Spain and Argentina performing well in terms of disrupting possession, while being closer to average in terms of conceding shots from counter-pressing possessions.

Many words have been and will continue to be expounded on Germany’s woes at this tournament, but their figures here merit a mention before turning to teams still in the World Cup. They were below average in terms of regaining possession, with a particularly woeful and damaging performance against Mexico. Toni Kroos was the thankless leader of Germany’s counter-press with Mats Hummels and Jérôme Boateng being the next most frequent counter-pressers, which seems shall we say sub-optimal and flies in the face of how other counter-pressing teams profile. Furthermore, 15% of the possessions they counter-pressed ended with their opponent taking a shot, tied for fourth-worst in the tournament. They were also tied for fourth-worst in terms of the number of shots conceded from counter-pressing possessions. Such figures are likely flattering given some of the counter-attacking opportunities wasted by Mexico and South Korea. Vulnerability isn’t a word usually associated with the German national team but that is what they were – yes, they were somewhat unfortunate to score only two goals while finishing bottom of their group but their struggles in defensive transition could have been even more disastrous against better opposition in the knock-out stages. Winning the World Cup is obviously Brazil’s priority but if they were ever going to avenge the 7-1 loss in 2014, then a last-16 clash against this German team would have been their chance.

Quarter-final notes

Uruguay vs France: neither team has been committed to counter-pressing during the tournament so far. When Uruguay have pressed, they haven’t typically regained possession quickly but they have been excellent at slowing down attacks and preventing shots. France have matched them in terms of shot-suppression, which suggests that the decisive moments will come from other angles.

Brazil vs Belgium: Brazil have sought to counter-press their opponents in their most crucial and more-even matches (vs Switzerland and Mexico), so may well aim for a similar strategy against Belgium. Discounting Belgium’s encounter with England (sorry Adnan) suggests they will also seek to put pressure on Brazil during transitions, which indicates the match will be keenly contested when possession is lost. Both teams looked somewhat vulnerable to counter-attacks during their round-of-16 matches and with each of them containing some of the best counter-attacking passers and finishers around, shutting-down or exploiting such opportunities could be key.

Sweden vs England: Sweden sit at the low end of the scale in terms of counter-pressing and have applied such tactics consistently across the tournament. Given their success across qualifying and at this tournament, we shouldn’t expect a deviation from this at the quarter-final stage. England employed an aggressive counter-press against Tunisia in their opener, followed by a drop off against Panama after running up an early lead, while the Belgium game is best forgotten (sorry Adnan). Their match against Colombia saw a return to more of a counter-pressing style, although not at the same level as against Tunisia. In summary, England have been more aggressive during transitions when they’ve had the most riding on their matches and have dominated possession, so expect them to utilize such tactics against Sweden.

Russia vs Croatia: the hosts have generally employed a less aggressive counter-press than their peers, with their lowest rate coming against Spain in the round-of-16. It will be interesting to see if they alter their approach against another technically gifted side. Croatia have counter-pressed at an average rate and at a quite consistent level, so it will be intriguing to see if they contest turn-overs or invite Russia onto them more with the intention of drawing them out of their defensive third.

Pressing has been a perhaps surprising feature of this World Cup and contributed to the narratives around several teams, whether they be success stories or futile failures. Come the 15th of July, we’ll see how much of a role it has played in the denouement of this wonderful World Cup.

Using Pressure to Evaluate Centre Backs

Originally published on StatsBomb.

Analysing centre backs is a subject likely to provoke either a shrug or a wistful smile from an analytics practitioner. To varying degrees, there are numbers and metrics aplenty for other positions but in public analytics at least, development has been limited and a genuine track record of successful application is yet to be found. If centre back analysis is the holy grail of public football analytics, then the search thus far has been more Monty Python than Indiana Jones.

One of the major issues with centre back analysis is that positioning isn’t measured directly by on-ball event data and any casual football watcher can tell you that positioning is a huge part of the defensive art. Tracking data would be the ideal means to assess positioning but it comes at a high-cost both computationally and technically, while having a much smaller coverage in terms of leagues than simpler event data provision.

StatsBomb’s new pressure event data serves as a bridge between the traditional on-ball event data and the detailed information provided by tracking data, offering a new prism to investigate the style and effectiveness of centre backs. While it won’t provide information on what a defender is up to when he is not in the immediate vicinity of the ball, it does provide extra information on how they go about their task.

Starting at the basic counting level, centre backs averaged six pressure actions per ninety minutes in the Premier League last season. Tackles and interceptions clock in at 0.8 and 1.3 per 90 respectively, which immediately illustrates that pressure provides a great deal more information to chew on when analysing more ‘proactive’ defending. I’m classing clearances and blocking shots as ‘reactive’ given they mostly take place in the penalty area and are more-directly driven by the opponent, while aerial duels are a slightly different aspect of defending that I’m going to ignore for the purposes of this analysis.

The figure below maps out where these defensive actions occur on the pitch and is split between left and right centre backs. Pressure actions typically occur in wider areas in the immediate vicinity of the penalty area, with another peak in pressure just inside the top corner of the 18-yard box. This suggests that centre backs don’t engage too high up the pitch in terms of pressure and are generally moving out towards the flanks to engage opponents in a dangerous position and either slow-down an attack, cut down an attackers options or directly contest possession.

DefensiveMaps.png

Maps illustrating the location of pressure actions, interceptions and tackles by centre backs in the 2017/18 EPL season. Top row is for left-sided centre backs and the bottom row is for right-sided centre backs.

The location of pressure actions is somewhat similar to the picture for interceptions, although the shape of the latter is less well-defined and tends to extend higher up the pitch. Tackles peak in the same zone just outside the top corners of the penalty area but are also less spatially distinct. Tackles also peak next to the edge of the pitch, a feature that is less distinct in the pressure and interception maps.

Partners in Crime

The number of pressure actions a centre back accumulates during a match will be driven by their own personal inclinations and role within the team, as well as the peculiarities of a given match and season e.g. the tactics of their own team and the opposition or the number of dangerous opportunities their opponent creates. The figure below explores this by plotting each individual centre back’s pressure actions per ninety minutes against their team name. The team axis is sorted by the average number of pressure actions the centre backs on each team make over the season.

CB_Pressure_Actions_per90

Pressure actions per 90 minutes by centre backs in the 2017/18 EPL season (minimum 900 minutes played) by team. Team axis is sorted by the weighted average number of pressure actions the centre backs on each team make over the season.

At the top end of the scale, we see Arsenal and Chelsea, two teams that regularly played a back-three over the past season. Nacho Monreal and César Azpilicueta led the league in pressure actions per ninety minutes by a fair distance and it appears the additional cover provided by playing in a back-three and their natural instincts developed as full backs meant they were frequently putting their opponents under pressure. Manchester United top the list in terms of those predominantly playing with two centre backs, with all of their centre backs applying pressure at similar rates.

At the other end of the scale, Brighton and Leicester’s centre backs appear to favour staying at home in general. Both though are clear examples of there being an obvious split between the number of pressure actions by the primary centre backs on a team, with one being more aggressive while the other presumably holds their position and plays a covering role. This division of roles is perhaps most clearly demonstrated by Chelsea’s centre backs, with Azpilicueta and Antonio Rüdiger as the side centre backs being more proactive than their counter-part in the central defensive slot (Cahill or Christensen).

Liverpool’s improved defensive performance over the course of the season has been attributed to a range of factors, with the signing of Virgil Van Dijk for a world-record fee garnering much of the credit. Intriguingly, his addition to the Liverpool backline has seemingly offered a significant contrast to the club’s incumbents, who all favoured a slightly greater than average number of pressure actions. Furthermore, Van Dijk ranked towards the bottom of the list in terms of pressure actions for Southampton (4.5 per 90) as well, with his figure for Liverpool (3.7 per 90) representing a small absolute decline. As an aside, Van Dijk brings a lot to the table in terms of heading skills, where he ranks highly for both total and successful aerial duels, so he is still an active presence in this aspect, while being a low-event player in others.

Centre backs are often referred to as a partnership and the above illustrates how defensive units often setup to complement each others skill sets and attempt to become greater than the sum of their parts.

The Thompson Triangle

Mark Thompson has led the way in terms of public analytics work on centre backs and has advocated for stylistic-driven evaluations as the primary means of analysis, which can then be built on with more traditional scouting. Pressure actions add another string to this particular bow and the figure below contrasts the three proactive defensive actions discussed earlier. Players in different segments of the triangle are biased towards certain actions, with those in the corners being more strongly inclined towards one action over the other two.

CB-TernaryGraph

Comparison of player tendencies in terms of ‘proactive’ defensive actions in the 2017/18 EPL season (minimum 900 minutes played). Apologies for triggering any flashbacks to chemistry classes. Click figure to open in new window.

There is a lot to pour over in the figure, so I’ll focus on defenders who are most inclined towards pressure actions. One clear theme is that such centre backs frequently featured on the sides of a back-three. Ryan Shawcross is unusual in this aspect given he was generally the middle centre back in Stoke’s back-three, as well as the right centre back in a back four. Ciaran Clark at Newcastle and Kevin Long at Burnley are the only players who featured mostly as one of two centre backs, with their partner adopting a more reserved role.

The additional cover provided by a back-three system and the frequent requirement for the player on the flanks to pull wide and cover in behind their wing-back seemingly plays a large part in determining the profile of centre backs. This illustrates the importance of considering team setup in determining a defenders profile and should feed into any recruitment process alongside their individual inclinations.

The analysis presented provides descriptive metrics and illustrations of the roles played by centre backs and is very much a first look at this new data. While we can’t gain definitive information on positioning without constant tracking of a player, the pressure event data provides a new lens to evaluate centre backs and significantly increases the number of defensive actions that can be evaluated further. Armed with such information, these profiles can be built upon with further data-driven analysis and combined with video and in-person scouting to build a well-rounded profile on the potential fit of a player.

Now all we need is a shrubbery.

Measuring counter-pressing

Originally published on StatsBomb.

The concept of pressing has existed in football for decades but its profile has been increasingly raised over recent years due to its successful application by numerous teams. Jürgen Klopp and Pep Guardiola in particular have received acclaim across their careers, with pressing seen as a vital component of their success. There are numerous other recent examples, such as the rise of Atlético Madrid, Tottenham Hotspur and Napoli under Diego Simeone, Mauricio Pochettino and Maurizio Sarri respectively.

Alongside this rise, public analytics has sought to quantify pressing through various metrics. Perhaps the most notable and widely-used example was ‘passes per defensive action’ or PPDA, which was established by Colin Trainor and first came to prominence on this very website. Anecdotally, PPDA found its way inside clubs and serves as an example of public analytics penetrating the private confines of football. Various metrics have also examined pressing through the prism of ‘possessions’, which Michael Caley has put to effective use on numerous occasions. Over the past year, I sought to illustrate pressing by quantifying a team’s ability to disrupt pass completion. While this was built on some relatively complex numerical modelling, it did provide what I thought was a nice visual representation of the effectiveness of a team’s pressing.

While the above metrics and others have their merits, they tend to ignore that pressing can take several forms and are biased towards the outcome, rather than the actual process. The one public example that side-steps many of these problems is the incredible work by the Anfield Index team through their manual collection of Liverpool’s pressing over the past few seasons but this has understandably been limited to one team.

Step-forward the new pressure event data supplied by StatsBomb Services. This new data is an event that is triggered when a player is within a five-yard radius of an opponent in possession. The radius varies as errors by the opponent would prove more costly, with a maximum range of ten-yards that is usually associated with goalkeepers under pressure. As well as logging the players involved in the pressure event and its location, the duration of the event is also collected.

The data provides an opportunity to explore pressing in greater detail than ever before. Different teams use different triggers to instigate their press, which can now be isolated and quantified. Efficiency and success can be separated from the pressing process in a number of ways at both the team and player-level. Such tools can be used in team-evaluation, opposition scouting and player recruitment.

One such application of the new data is to explore gegenpressing or counter-pressing, which is the process where a team presses the opposition immediately after losing possession. The initial aim of counter-pressing is to disrupt the opponent’s counter-attack, which can be a significant danger during the transition phase from attack-to-defence when a team is more defensively-unstable. Ideally possession is quickly won back from the opponent, with some teams seeking to exploit such situations to attack quickly upon regaining possession. Five seconds is often used as a cut-off for the period where pressure on the opposition is most intensely applied during the counter-press.

The exciting new dimension provided by StatsBomb’s new pressure data is that the definition of counter-pressing you would find in a coaching manual can be directly drawn from the data i.e. a team applies pressure to their opponent following a change in possession. The frequency at which counter-pressing occurs can be quantified and then we can develop various metrics to examine the success or failure of this process. Furthermore, we can analyse counter-pressing at the player-level, which has been out-of-reach previously.

The figure below illustrates where on the pitch counter-pressing occurs based on data from 177 matches from the Premier League this past season. The pitch is split into six horizontal zones and is orientated so that the team out-of-possession is playing from left-to-right. The colouring on the pitch shows the proportion of open-play possessions starting in each zone where pressure is applied within five seconds of a new possession.

AvgCounterPressMap.png

The figure illustrates that pressure is most commonly applied on possessions starting in the midfield zones, with marginally more pressure in the opposition half. Possessions beginning in the highest zone up the pitch come under less pressure, which is likely driven by the lower density of players in this zone on average. Very few possessions actually begin in the deepest zone and a smaller proportion of them come under pressure quickly than those in midfield.

From a tactical perspective, pressing is generally reserved for areas outside of a team’s own defensive third. The exact boundary will vary but for the following analysis, I have only considered possessions starting higher up the pitch, as denoted by the counter-pressing line in the previous figure.

In the figures below, the proportion of possessions in the counter-pressing zones where pressure is applied within five seconds is referred to as the ‘counter-pressing fraction’. In the sample of matches from the Premier League this season, a little under half (0.47) of open-play possessions come under pressure from their opponent within five seconds. At the top of the counter-pressing rankings, we see Manchester City, Tottenham Hotspur and Liverpool, which is unsurprising given the reputations of their managers. At the bottom end of the scale, we find a collection of teams that have mostly been overseen by British managers who are more-known for a deep-defensive line.

Team_CounterPressFraction

On the right-hand figure above, the strong association between counter-pressing and possession is illustrated, with the two showing a high correlation coefficient of 0.86 in this aggregated sample. Interpreting causality here is somewhat problematic given the likely circular relationship between the two parameters; teams that dominate possession may have more energy to press intensively, leading to a greater counter-pressing fraction, which would lead to them winning possession back more quickly, which will potentially increase their possession share and so on. The correlation is weaker for individual matches (0.36), which hints at some greater complexity and is something that can be returned to at a later date.

Perhaps the most interesting finding in the above figures is Burnley’s high counter-pressing fraction. The majority of analysis on Burnley has focused on their defensive structure within their own box and how that affects their defensive performance in relation to expected goals. The figure illustrates that Burnley employ a relatively aggressive counter-press, especially in relation to their possession share.

Examining Burnley’s counter-pressing game in more detail reveals that they counter-press 18 possessions per game, which is above average and only slightly lower than Manchester City. However, they only actually regain possession within five seconds 2.5 times per game, which falls short of what you might expect on average and falls below their counter-pressing peers. In terms of the ratio between their counter-pressing regains and total counter-pressing possessions, they sit 17th on 14%.

Burnley’s counter-press is the fourth least-effective at limiting shots, with 13% of such possessions ending with them conceding a shot compared to the average rate of 10%. However, one thing in their favour is that these possessions are typically around the league average in terms of their length and speed of attack, which will allow Burnley to regain their vaunted defensive organisation prior to conceding such shots.

The more dominant discourse around pressing is as an attacking rather than defensive weapon, so narratives are often formed around teams that regularly win back the ball through pressing and use this to generate fast attacks e.g. Liverpool and Tottenham Hotspur. As a result, a team like Burnley who seemingly employ counter-pressing as a defence-first tactic to prevent counter-attacks and slow attacking progress may be overlooked.

Burnley’s manager, Sean Dyche, has typically been lumped-in with the tactical stylings of the perennially-employed British managers who aren’t generally associated with pressing tactics. Dyche was reportedly most impressed by the pressing game employed by Guardiola’s Barcelona and he has seemingly implemented some of these ideas at Burnley. He has instilled an approach that combines counter-pressing and a low-block with numbers behind the ball, which is a neat trick to pull-off; Diego Simeone and Atlético Madrid are perhaps the more apt comparison given such traits.

The above analysis illustrates the ability of StatsBomb’s new pressure event data to illuminate an important aspect of the modern game. Furthermore, it is able to do this in a manner that directly translates tactical principles, separating underlying process and outcome, which is a giant step-forward for analytics. It also led to an analysis discussing the similarity between Guardiola’s legendary Barcelona team and Sean Dyche’s Burnley, which was probably unexpected to say the least.

This is just a taster of what is possible with StatsBomb’s new data. There’s more information in this presentation from the StatsBomb launch event and you can expect more analysis to appear over the summer and beyond.

Liverpool and I

While I probably watched Liverpool play before then, the first match I remember watching was on the 4th January 1994, when a nine-year-old me saw them come back from three goals down, which would become something of a theme. As is the want of memory, the events that leave an indelible mark are the ones that stand-out; my first actual football memory is Paul Bodin missing that penalty and not really understanding the scale of the disappointment. Turned out Wales’ last World Cup match was in 1958 when some no-mark seventeen-year-old called Edson Arantes do Nascimento scored his first international goal and knocked them out in the quarter-final.

Other early memories include one of God’s defining miracles, with a hat-trick notched up in four minutes and thirty three seconds and learning about player aging curves when I realised that the slow yet classy guy in midfield used to be one of the most devastating and exciting wide-players the game had ever seen. My first match at Anfield was Ian Rush’s last there in a red shirt, while subsequent visits took in thrilling cup matches under the gaze of King Kenny and the best live sporting experience of my life as I bounced out of Anfield full of hope in April 2014.

While a league title has proved elusive during my supporting life, Europe has provided the greatest thrills, with tomorrow marking a third European Cup Final to go along with two finals in the junior competition. A European Cup Final once every eight years on average, with all three in the last fourteen years is pretty good going for a non-super club, albeit one with significant resources.

Real Madrid are clearly going to be a tough nut to crack, with Five Thirty Eight, Club Elo and Euro Club Index all ranking them as the second best team around. The same systems have Liverpool as the fifth, seventh and eleventh best, so under-dogs with a good chance at glory overall.

According to Club Elo, the 2018 edition of Liverpool will be the best to contest a European Cup Final this century but on the flip-side, Real Madrid are stronger than either of the AC Milan teams that they faced in 2005 and 2007. Despite this, Liverpool are given a slightly better shot at taking home Old Big Ears than they had in 2005, as the gap between them and their opponents is narrower. The strides that the team made under Rafa between the 2005 and 2007 finals meant that the latter was contested by two equal teams.

Liverpool should evidently be approaching the final with optimism and further evidence of this is illustrated in the figure below, which shows the top-fifty teams by non-penalty expected goal difference in the past eight Premier League seasons. The current incarnation of Liverpool sit fifth and would usually be well-positioned to seriously challenge for the title. As the figure also illustrates, the scale of Manchester City’s dominance in their incredible season is well-warranted.

EPL-8-seasons-xGD.png

Top-fifty teams by non-penalty expected goal difference over the past eight Premier League seasons. Liverpool are highlighted in red, with the 17/18 season marked by the star marker. Data via Opta.

Liverpool’s stride forward under Klopp this past season has taken them beyond the 13/14 and 12/13 incarnations in terms of their underlying numbers. In retrospect, Rodgers’ first season was quietly impressive even if it wasn’t reflected in the table and it set the platform for the title challenge the following season.

Compared to those Luis Suárez-infused 12/13 and 13/14 seasons, the attacking output this past season is slightly ahead, with the team sitting sixth in the eight-season sample, which is their best over the period. Including penalties would take the 13/14 vintage beyond the latest incarnation, with the former scoring ten from the twelve (!) awarded, while 17/18 saw only three awarded (two scored).

The main difference between the current incarnation though is on the defensive end, with the team having the fifth best record in terms of non-penalty expected goals conceded this past season in the eight-year sample. The 13/14 season’s defence was the seventh worst by the club in this eight-year period and they lay thirty-fourth overall. These contrasting records equate to an eight non-penalty expected goal swing in their defensive performance.

While the exhilarating attacking intent of this Liverpool side is well-established, they are up against another attacking heavyweight; could it be that the defensive side of the game is the most decisive? The second half of this season is especially encouraging on this front, with improvements in both expected and actual performance. This period represents the sixth best half season over these eight-seasons (out of a total of 320) and a three-goal swing compared to the first half of the season. This was slightly offset by a reduction in attacking output of two non-penalty expected goals but the overall story is one of improvement.

The loss of Coutinho, addition of van Dijk and employing a keeper with hands (edit 2203 26/05/18: well at least he gets his hands to it usually) between the sticks is a clear demarcation in Liverpool’s season and it is this period that has seen the thrilling run to the European Cup Final. The improved balance between attack and defence bodes well and I can’t wait to see what this team can do on the biggest stage in club football.

Allez, Allez, Allez!

On Burnley and expected goals

Originally published on StatsBomb.

Expected goals has found itself outside the confines of the analytics community this season, which has brought renewed questions regarding its flaws, foibles and failures. The poster-child for expected goals flaws has been Burnley and their over-performing defence, even finding themselves in the New York Times courtesy of Rory Smith. Smith’s article is a fine piece blending comments and insights from the analytics community and Sean Dyche himself.

Smith quotes Dyche describing Burnley’s game-plan when defending:

The way it is designed is to put a player in a position that it is statistically, visually and from experience, harder to score from.

Several analyses have dug deeper into Burnley’s defence last season, including an excellent piece by Mark Thompson for StatsBomb. In his article, Mark used data from Stratagem to describe how Burnley put more men between the shooter and the goal than their peers, which may go some way to explaining their over-performance compared with expected goals.

Further work on the EightyFivePoints blog quantified how the number of intervening defenders affected chance quality. The author found that when comparing an expected goal model with and without information on the number of intervening defenders and a rating for defensive pressure, Burnley saw the largest absolute difference between these models (approximately 4 goals over the season).

If there is a quibble with Smith’s article it is that it mainly focuses on this season, which was only 12 games old at the time of writing. Much can happen in small samples where variance often reigns, so perhaps expanding the analysis to more seasons would be prudent.

The table below summarises Burnley’s goals and expected goals conceded over the past three and a bit seasons.

xG_Table

Burnley’s non-penalty expected goals and goals conceded over the past four seasons. Figures for first 13 games of 2017/18 season. Data via Opta.

Each season is marked by over-performance, with fewer goals conceded than expected. This ranges from 5 goals last season to a scarcely believable 15 goals during their promotion season in the Championship.

The above table and cited articles paint a picture of a team that has developed a game-plan that somewhat flummoxes expected goals and gains Burnley an edge when defending. However, if we dig a little deeper, the story isn’t quite as neat as would perhaps be expected.

Below are cumulative totals for goals and expected goals as well as the cumulative difference over each season.

Burnley_xG_Timeline_Full

Burnley’s cumulative non-penalty goals and expected goals conceded (left) and the cumulative difference between them (right) over the past four seasons. Figures for first 13 games of 2017/18 season. Data via Opta.

In their 2014/15 season, Burnley were actually conceding more goals than expected for the majority of the season until a run of clean sheets at the end of the season saw them out-perform expected goals. After the 12 game mark in their Championship season, they steadily out-performed expected goals, yielding a huge over-performance. This continued into their 2016/17 season in the Premier League over the first 10 games where they conceded only 12 goals compared to an expected 19. However, over the remainder of the season, they slightly under-performed as they conceded 39 goals compared with 36 expected goals.

The above illustrates that Burnley’s over-performance in previous Premier League seasons is actually driven by just a handful of games, rather than a systematic edge when examining expected goals.

This leaves us needing to put a lot of weight on their Championship season if we’re going to believe that Burnley are magical when it comes to expected goals. While the descriptive and predictive qualities of expected goals in the Premier League is well-established, there is less supporting evidence for the Championship. Consequently it may be wise to take their Championship figures with a few grains of salt.

This season and last has seen Burnley get off to hot starts, with what looks like a stone-cold classic example of regression to the mean last season. If we ignore positive variance for a moment, perhaps their opponents got wise to their defensive tactics and adapted last season but then you have to assume that they’ve either forgotten the lessons learned this season or Dyche has instigated a new and improved game-plan.

The cumulative timelines paint a different picture to the season aggregated numbers, which might lead us to conclude that Burnley’s tactics don’t give them quite the edge that we’ve been led to believe. In truth we’ll never be able to pin down exactly how much is positive variance and how much is driven by their game-plan.

However, we can state that given our knowledge of the greater predictive qualities of expected goals when compared to actual goals, we would expect Burnley’s goals against column to be closer to their expected rate (1.3 goals per game) than their current rate (0.6 goals per game) over the rest of the season.

Time will tell.

What has happened to the Klopp press?

Originally published on StatsBomb.

When asked how his Liverpool team would play by the media horde who greeted his unveiling as manager two years ago, Jürgen Klopp responded:

We will conquer the ball, yeah, each fucking time! We will chase the ball, we will run more, fight more.

The above is a neat synopsis of Klopp’s preferred style of play, which focuses on pressing the opponent after losing the ball and quickly transitioning into attack. It is a tactic that he successfully deployed at Borussia Dortmund and one that he has employed regularly at Liverpool.

However, a noticeable aspect of the new season has been Liverpool seemingly employing a less feverish press. The Anfield Index Under Pressure Podcast led the way with their analysis, which was followed by The Times’ Jonathan Northcroft writing about it here and Sam McGuire for Football Whispers.

Liverpool’s pass disruption map for the past three seasons is shown below. Red signifies more disruption (greater pressure), while blue indicates less disruption (less pressure). In the 2015/16 and 2016/17 seasons, the team pressed effectively high up the pitch but that has slid so far this season to a significant extent. There is some disruption in the midfield zone but at a lower level than previously.

LFC_dxP.png

Liverpool’s zonal pass completion disruption across the past three seasons. Teams are attacking from left-to-right, so defensive zones are to the left of each plot. Data via Opta.

The above numbers are corroborated by the length of Liverpool’s opponent possessions increasing by approximately 10% this season compared to the rest of Klopp’s reign. Their opponents so far this season have an average possession length of 6.5 seconds, which is lower than the league average but contrasts strongly with the previous figures that have been among the shortest in the league.

Examining their pass disruption figures game-by-game reveals further the reduced pressure that Liverpool are putting on their opponents. During 2015/16 and 2016/17, their average disruption value was around -2.5%, which they’ve only surpassed once in Premier League matches this season, with the average standing at -0.66%.

LFC-xP-17-18

Liverpool’s game-by-game pass completion disruption for 2017/18 English Premier League season. Figures are calculated for zones above Opta x-coordinates greater than 40. Data via Opta.

The Leicester match is the major outlier and examining their passing further indicates that the high pass disruption was a consequence of them attempting a lot of failed long passes. This is a common response to Liverpool’s press as teams go long to bypass the pressure.

Liverpool’s diminished press is likely a deliberate tactic that is driven by the added Champions League matches the team has faced so far this season. The slightly worrisome aspect of this tactical shift is that Liverpool’s defensive numbers have taken a hit.

In open-play, Liverpool’s expected goals against figure is 0.81 per game, which is up from 0.62 last season. Furthermore, their expected goals per shot has risen to 0.13 from 0.11 in open-play. To add further defensive misery, Liverpool’s set-piece woes (specifically corners) have actually got worse this season. The team currently sit eleventh in expected goals conceded this season, which is a fall from fifth last year.

This decline in underlying defensive performance has at least been offset by a rise on the attacking side of 0.4 expected goals per game to 1.78 this season. Overall, their expected goal difference of 0.79 this season almost exactly matches the 0.81 of last season.

Liverpool’s major problem last season was their soft under-belly but they were often able to count on their pressing game denying their opponents opportunities to exploit it. What seems to be happening this season is that the deficiencies at the back are being exploited more with the reduced pressure ahead of them.

With the season still being relatively fresh, the alarm bells shouldn’t be ringing too loudly but there is at least cause for concern in the numbers. As ever, the delicate balancing act between maximising the sides attacking output while protecting the defense is the key.

Klopp will be searching for home-grown solutions in the near-term and a return to the familiar pressing game may be one avenue. Given the competition at the top of the table, he’ll need to find a solution sooner rather than later, lest they be left behind.

Under pressure

Originally published on StatsBomb.

Models that attempt to measure passing ability have been around for several years, with Devin Pleuler’s 2012 study being the first that I recall seeing publicly. More models have sprung up in the past year, including efforts by Paul Riley, Neil Charles and StatsBomb Services. These models aim to calculate the probability of a pass being completed using various inputs about the start and end location of the pass, the length of the pass, the angle of it, as well as whether it is played with the head or foot.

Most applications have analysed the outputs from such models from a player passing skill perspective but they can also be applied at the team level to glean insights. Passing is the primary means of constructing attacks, so perhaps examining how a defense disrupts passing could prove enlightening?

In the figure below, I’ve used a pass probability model (see end of post for details and code) to estimate the difficulty in completing a pass and then compared this to the actual passing outcomes at a team-level. This provides a global measure of how much a team disrupts their opponents passing. We see the Premier League’s main pressing teams with the greatest disruption, through to the barely corporeal form represented by Sunderland.

Team_PCDgraph

Pass completion disruption for the 2016/17 English Premier League season. Disruption is defined as actual pass completion percentage minus expected pass completion percentage. Negative values means opponent’s complete fewer passes than expected. Data via Opta.

The next step is to break this down by pitch location, which is shown in the figure below where the pitch has been broken into five bands with pass completion disruption calculated for each. The teams are ordered from most-to-least disruptive.

PressureMap

Zonal pass completion disruption for 2016/17 English Premier League season. Teams are attacking from left-to-right, so defensive zones are to the left of each plot. Data via Opta.

We see Manchester City and Spurs disrupt their opponents passing across the entire pitch, with Spurs’ disruption skewed somewhat higher. Liverpool dominate in the midfield zones but offer little disruption in their deepest-defensive zone, suggesting that once a team breaks through the press, they have time and/or space close to goal; a familiar refrain when discussing Liverpool’s defense.

Chelsea offer an interesting contrast with the high-pressing teams, with their disruption gradually increasing as their opponents inch closer to their goal. What stands out is their defensive zone sees the greatest disruption (-2.8%), which illustrates that they are highly disruptive where it most counts.

The antithesis of Chelsea is Bournemouth who put together an average amount of disruption higher up the pitch but are extremely accommodating in their defensive zones (+4.5% in their deepest-defensive zone). Sunderland place their opponents under limited pressure in all zones aside from their deepest-defensive zone where they are fairly average in terms of disruption.

The above offers a glimpse of the defensive processes and outcomes at the team level, which can be used to improve performance or identify weaknesses to exploit. Folding such approaches into pre-game routines could quickly and easily supplement video scouting.

Appendix: Pass probability model

For this post, I built two different passing models; the first used Logistic Regression and the second used Random Forests. The code for each model is available here and here.

Below is a comparison between the two, which compares expected success rates with actual success rates on out-of-sample test data.

Actual_vs_Expected

Actual versus expected pass success for two different models. Data via Opta.

The Random Forest method performs better than the Logistic Regression model, particularly for low probability passes. This result is confirmed when examining the Receiver Operating Characteristics (ROC) curves in the figure below. The Area Under the Curve (AUC) for the Random Forest model is 87%, while the Logistic Regression AUC is 81%.

xPass_AUC

Receiver Operating Characteristics (ROC) curves for the two different passing models. Data via Opta.

Given the better performance of the Random Forest model, I used this in the analysis in the main article.

Liverpool 2017/18 season preview

Originally published on StatsBomb.

Liverpool enter the season with aspirations of challenging for the title after an at times hugely promising and exciting first full season under Jürgen Klopp. The prospect of European adventures returning on Tuesday or Wednesday nights is tantalizing close providing they negotiate their Champions League qualifying round.

The story so far

Liverpool’s tally of 76 points last season was their joint-third best tally over the last decade and only their second top-four finish since the Benitez years. In fact, after a run of four top-four finishes, Liverpool haven’t registered back-to-back Champions League qualifications since Rafa left and have on average finished sixth during that time with 65 points on the board.

With the above in mind, it’s tempting to view a season of consolidation as the priority for the coming season, alongside beginning to re-establish the team as a European force. Liverpool’s underlying performance last season is encouraging, with their goal return reasonably in-line with expectation and their expected goal difference placing them well in contention for a title push.

Drilling further into their expected goal numbers, sees a team that experienced fluctuating under-lying performance over the course of the season with a significant decline once 2017 was rung in. The graphic below illustrates this alongside a longer-term outlook encompassing the past five seasons.

LFC_xG_TimeLine.png

Rolling 19-game average expected goal timeline over the past five seasons. Grey vertical lines denote new season.

The heights of 2016/17 are close to those of the Suárez-powered team under Rodgers, while the low-point is more in-line with Klopp’s early tenure at the club. The past season thus illustrated that the team was capable of title-contending performances at times but also switched to a team competing for the fourth-place trophy at best.

Upping the pace

Closer examination of the downturn in performance using my ‘team strategy analysis‘ shows a drying up of shot generation via high-quality chances born of fast-paced attacks from deep and after midfield-transitions.

Sadio Mané was evidently missed due to AFCON duties and injury over the latter half of the season and this is borne out by the numbers. According to my model, he was second best in the EPL (0.11 per 90) in terms of xG-contribution (the sum of expected goals and assists) from fast-paced attacks following a midfield-transition. For fast-attacks from deep, he ranked sixth for xG-contribution (0.12 per 90).

Thankfully, Mohamed Salah, the club’s major acquisition so far, brings complementary qualities to the table and adds much-needed depth to the wide-forward ranks. James Yorke of this parish has already praised the signing earlier this summer and my only addition is that Salah showed up quite highly for xG-contribution (0.07 per 90, ranking eleventh in Serie A) for fast-paced attacks following a midfield-transition. The addition of Salah improves what was already a healthy front-line attack.

Defensive issues

According to the Objective Football website run by Benjamin Pugsley, Liverpool conceded just 8.1 non-penalty shots per game, ranking second over the past eight seasons behind a Pep-infused Manchester City last year. Shots-on-target conceded (3.0 per game) told a similar story, ranking joint-sixth over the same period. However, they combined these extraordinary shot-suppression numbers with the highest expected goals per shot in the league (0.11), which is the worst value I have over the past five seasons. When Liverpool conceded shots, they were of high quality, which ultimately saw them sit fifth in terms of expected goals against last season.

Klopp’s tactical system deserves credit for melding a highly exciting attack with strong defensive aspects in terms of shot-suppression. The optimistic take here is that tweaks and a greater familiarity with his counter-pressing tactics could bring about improvements in shot quality conceded, thereby seeing better defensive numbers. It’s worth noting the period during November and December 2016 when their expected goals against was the lowest it has been consistently over a 19-game span in the past five seasons, so the current squad is capable of sustained excellence in this realm.

The pursuit of Virgil van Dijk does suggest that the club are aiming to recruit a new starting centre-back. That saga remains running at the time of writing as the world waits to find out just how costly a single ice cream can be. Centre-back depth is an issue that needs to be rectified; Lucas Leiva made six appearances as a centre-back last term is all the evidence needed for that statement.

The other aspect of Liverpool’s defense that could improve is in the goalkeeping stakes. From a pure shot-stopping perspective, Karius has the best pedigree; in my goalkeeper shot-stopping analysis, Karius came 31st across the data-set with a rating of 91%, which is a pretty decent indication that he is an above-average shot-stopper. Mignolet fared much worse with a ranking of just 25%, which puts him at best as an average shot-stopper during his Liverpool career to date. I haven’t looked at numbers for the Championship but Mark Taylor’s numbers for Ward at Huddersfield were not encouraging. Playing Karius would be a bold move by Klopp given his limited exposure to English football thus far but Mignolet doesn’t provide much confidence either -personally, I would go with Karius.

Title talk

If I’ve learnt anything while sifting through the data for this preview, it’s that Manchester City should be strong favourites for the title this coming season.

Can Liverpool challenge them, while also competing in Europe? At present, I’d side with no given the depth issues of last season have yet to be addressed and the remaining questions marks in terms of the defense.

Liverpool’s other transfer saga involving Naby Keita could be a game-changer given that he could have a transformative impact on the team’s midfield but the likelihood of him signing appears to be receding by the day. Midfield depth is also potentially an issue unless Klopp is happy to rely on youth to cover midfield absentees over the season.

With potentially five teams in the Champions League group stages, progress to the latter rounds could have a strong bearing on league form post-Christmas. Six into four is likely the maths heading into the new season and Liverpool should be well in the mix.

Prediction: Third We’re gonna win the league

More thinking about goalkeepers

Following my previous article on the shot-stopping ability of goalkeepers, Mike Goodman posed an interesting question on Twitter:

This is certainly not an annoying question and I tend to think that such questions should be encouraged in the analytics community. Greater discussion should stimulate further work and enrich the community.

It certainly stands to reason and observation that goalkeepers can influence a strikers options and decision-making when shooting but extracting robust signals of such a skill may prove problematic.

To try and answer this question, I built a quick model to calculate the likelihood that a non-blocked shot would end up on target. It’s essentially the same model as in my previous post but for expected shots on target rather than goals. The idea behind the model is that goalkeepers who are able to ‘force’ shots off-target would have a net positive rating when subtracting actual shots on target from the expected rate.

When I looked at the results, two of the standout names were Gianluigi Buffon and Jan Oblak; Buffon is a legend of the game and up there with the best of all time, while Oblak is certainly well regarded, so not a bad start.

However, after delving a little deeper, dragons started appearing in the analysis.

In theory, goalkeepers influencing shot-on-target rates would do so for shots closer to goal as they would narrow the amount of goal they can aim for via their positioning. However, I found the exact opposite. Further investigation of the model workings pointed to the problem – the model showed significant biases depending on whether the shot was inside or outside the area.

This is shown below where actual and expected shot-on-target totals for each goalkeeper are compared. For shots inside the box, the model tends to under-predict, while the opposite is the case for outside the box shots. These two biases cancelled each other out when looking at the full aggregated numbers (the slope was 0.998 for total shots-on-target vs the expected rate).

Act_vs_Ex_SoT.png

Actual vs expected shots-on-target totals for goalkeepers considered in the analysis. Dashed line is the 1:1 line, while the solid line is the line of best fit. Left-hand plot is for shots inside the box, while the right-hand plot is for shots outside the box. Data via Opta.

The upshot of this was that goalkeepers performing well-above expectation were doing so due to shots from longer-range being off-target when compared to the expected rates for the model. I suspect that the lack of information on defensive pressure is skewing the results and introducing bias into the model.

Now when we think of Buffon and Oblak performing well, we recall that they play behind probably the two best defenses in Europe at Juventus and Atlético respectively. Rather than ascribing the over-performance to goalkeeping skill, the effect is likely driven by the defensive pressure applied by their team-mates and issues with the model.

Exploring model performance is something I’ve written about previously and I would also highly recommend this recent article by Garry Gelade on assessing expected goals. While the above is an unsatisfactory ending for the analysis, it does illustrate the importance of testing model output prior to presenting results and testing whether such results match with our theoretical expectations.

Knowing what questions analytics can and cannot answer is a pretty useful thing to know. Better luck next time hopefully.